Attenuators between a tube amp and the guitar speaker: some measurements and theory

[Mike Sulzer]

Any progress on a physical prototype?
The large capacitance is a challenge, though. Plus, isn't the need to be tethered to an external power supply for a proper NP electrolytic capacitor supply kind of a drag?
If you're willing to sacrifice some physical space, I think you can achieve this capacitance using much higher quality film capacitors)
For instance, doing a quick search, I you could build a bank of five 70µF/500V capacitors totaling an area of 152.5mm(L) X 57.5mm(W) X 50mm(H) for $20.95. Obviously, 500V is not needed and lower voltage/smaller caps are available, but I selected these were less expensive.

Thanks for checking up! That sounds like a bargain on film capacitors. I would want to determine how much current they can handle, but probably it is plenty.

I have had some hardware issues with my measurement system, and my new "Christmas present to myself" is almost here. Then I will check the measurements again.

I do not yet know how important the bias on the series electrolytics is. That is something to be tested.

The heaviest thing is the inductor for the low frequency resonance; that is, it would be good if that is not necessary. I suspect that the low frequency resonance itself might not be necessary, but the rising frequency response below 300 might be, and it might be achieved approximately (maybe "good enough") without the inductor, but there is testing and simulation to do there. It would provably require more capacitance, one reason for sticking with electrolytics.

[Gregg]

The heaviest thing is the inductor for the low frequency resonance; that is, it would be good if that is not necessary. I suspect that the low frequency resonance itself might not be necessary

The low frequency resonance is VERY important! Especially when kranked a tube will generate certain harmonics in this frequency region which if not properly loaded by this inductor otherwise won't be present. Even if you cut the fundamental during post processing you still will be able to hear those harmonics especially in the low and low mid range. You CAN'T add those later with an EQ if they are not initially present.

Concerning the capacitors I don't know why so many people are trying to discover the wheel. That reactive load has been built by many people and every aspect of it is described in detail over several forums. Bottom line is get a beefy motor start capacitor which can handle continuous high currents and forget about it.

I've built several of those reactive loads and they work very well. Also they measure as close to a real cab as you can get with those few parts.

This is an article about reactive load attenuators from Victor Kempf who is one of the AMT masterminds. Unfortunately it looks like the site doesn't load correctly in Google translate (at least on my PC) so maybe you'll need to copy/paste the text separately.

http://media.amtelectronics.com/gita...ciej-nagruzki/

There's also a more detailed discussion in their forum on the link below the article.

[Mike Sulzer]

The low frequency circuit is resonant. There are no harmonics generated at this resonance; it is too low. There are harmonics generated above the resonance where the low frequency circuit still has some effect, but the impedance is capacitive in this range.

I have not translated the text at your link, but I have looked at the schematic. As far as loading the amp goes, they have apparently missed the frequency dependent loss in the high frequency inductor. Maybe they have shown that this does not matter, but I doubt it. Why are they using an 8 ohm resistor for the voice coil resistance? 6.5 is more typical of guitar speakers. Possibly this is to compensate for the resistor divider they use to feed the speaker.

The idea is to load the amp properly, and then duplicate this voltage, scaled down, at the speaker. A resistor network such as they use only does this if the impedance looking back into it is significantly lower than the lowest impedance of the speaker over the necessary frequency range. Four ohms is not low enough.

If you turn on a sine wave at the resonance (low frequency), it draws current as limited by the voice coil resistance at first, but then the current drops well below the value it would have without the resonance. If the sine wave is turned off, energy is returned at the resonance frequency. Thus it stretches the signal. Especially with the bridge pickup, the level of the fundamental is quite low. So I am not convinced that this matters much, but I see it as something that needs to be determined.

EDIT: I forgot to say that the restive divider used in the Russian circuit can be replaced with a transformer. This does a better job of copying a reduced voltage from the load to the speaker. What you want is a "bridging" transformer, one that is considerably higher than 8 ohms looking into it. One way to do this is to (mis)use a so called line matching transformer, such as the Hammond 117 series. These are still available, believe it or not. It looks to me that a 25 volt model would do better than a 70 volt, but I have not tried this yet, and some thinking and computing is necessary.

The low frequency resonance is VERY important! Especially when kranked a tube will generate certain harmonics in this frequency region which if not properly loaded by this inductor otherwise won't be present. Even if you cut the fundamental during post processing you still will be able to hear those harmonics especially in the low and low mid range. You CAN'T add those later with an EQ if they are not initially present.

Concerning the capacitors I don't know why so many people are trying to discover the wheel. That reactive load has been built by many people and every aspect of it is described in detail over several forums. Bottom line is get a beefy motor start capacitor which can handle continuous high currents and forget about it.

I've built several of those reactive loads and they work very well. Also they measure as close to a real cab as you can get with those few parts.

This is an article about reactive load attenuators from Victor Kempf who is one of the AMT masterminds. Unfortunately it looks like the site doesn't load correctly in Google translate (at least on my PC) so maybe you'll need to copy/paste the text separately.

http://media.amtelectronics.com/gita...ciej-nagruzki/

There's also a more detailed discussion in their forum on the link below the article.

[dai h.]

re: Google translated article and the (I'm presuming) orange floating menu thing that obscures the article, (in Chrome (browser)) I tried "Inspect" (command--right click mouse on the page to see it come up), then deleting an element that says something like "this is fixed floating main menu" (then closing all the stuff that comes up after clicking the "Inspect" command (click "X" on the right hand side top)), and the orange floating menu thing was removed.

(In Firefox, there is a "Web Developer" command where (I assume) you can do the same thing (delete that element) but I'm not particularly computer savvy.)

[Chuck H]

Nice to see you dai.

[Helmholtz]

I will use this:
https://www.mouser.de/ProductDetail/...2fKzbD5w%3d%3d

Has to fit in my Power Break:

[Gregg]

As far as loading the amp goes, they have apparently missed the frequency dependent loss in the high frequency inductor.

For attenuator purposes maybe it's not necessary. In reactive load only circuits usually a 33-47Ohm resistor is used.

Below is an impedance measurement of a real 2xV30s cab (red) vs an unloaded HF inductor (yellow) vs 33Ohm loaded HF inductor (green). Inductor is 470uH.



You can hear the difference between loaded and unloaded but I don't think it's that great.

[Mike Sulzer]

As Helmholts pointed out, a better way is two inductors and resistors. This gets amplitude and phase right. Phase can matter because the harmonics generated by a nonlinear circuit depend on the shape of the waveform, and both the amplitude and phase contribute to the shape.

For attenuator purposes maybe it's not necessary. In reactive load only circuits usually a 33-47Ohm resistor is used.

Below is an impedance measurement of a real 2xV30s cab (red) vs an unloaded HF inductor (yellow) vs 33Ohm loaded HF inductor (green). Inductor is 470uH.

[ATTACH=CONFIG]51510[/ATTACH]

You can hear the difference between loaded and unloaded but I don't think it's that great.

[Gregg]

No doubt all this can be improved the question is whether for practical purposes the difference would justify the extra effort. Maybe it's one of those things that you have to try for yourself (including biasing the capacitors - seriously??).

[Helmholtz]

An alternative solution for the HF part of the impedance:
hfz.pdf

Needed something to be wired in parallel with the existing 1mH choke in the PB 100.

[Mike Sulzer]

No doubt all this can be improved the question is whether for practical purposes the difference would justify the extra effort. Maybe it's one of those things that you have to try for yourself (including biasing the capacitors - seriously??).

Right. The way I do that is to make something as good as I know how, and then try throwing out things that might not be needed. That includes things that you might consider necessary but have not actually shown to be so.

You do understand that NP electrolytics have to be biased and that self biasing might not be adequate in all cases? The page I posted above explains that well, but I figured this out independently. (Most explanations of now NPs work are seriously incomplete, and it took a while to find that one.)

[dai h.]

Nice to see you dai.

hi Chuck (mostly in "quiet mode" here)

[Chuck H]

Phase can matter because the harmonics generated by a nonlinear circuit depend on the shape of the waveform, and both the amplitude and phase contribute to the shape.

This^^^^^^^^^^^^^^^^^^^^

[Chuck H]

hi Chuck (mostly in "quiet mode" here)

Being old school, and so, used to seeing your posts, I totally missed that you posted on the "tubes that can handle 500V" thread. Then when I saw you here it rung a bell haven't had the pleasure of your input in awhile. Lurk as you like. Just glad you're still contributing whenever you see fit

[SoulFetish]

The real (the apparent resistance as a function of frequency) and imaginary (the apparent inductance as a function of frequency) parts are shown in the next attachment.


[ATTACH=CONFIG]51083[/ATTACH]
Now we see that the rise in impedance is a result of both inductance and increasing resistance. The latter is probably a result of eddy current losses in the metal parts.

Mike (this may be a dumb question), but when analyzing the effect on high frequencies, were you measuring the phase shift to differentiate between the real/imaginary components of the impedance?